Words like cleave and egregious have meanings that are completely opposite to other meanings of the same word. How did such a bizarre, confusing state of affairs ever develop?

I mean, I just can't work out how one and the same word could come to mean two distinct things that are completely opposite! Is it the same root? What did the root mean and how did it come to be used to mean the opposite?

  • 2
    This kind of word is known as an autoantonym. en.wikipedia.org/wiki/Auto-antonym – Urbycoz Jun 14 '11 at 15:23
  • 2
    Think of recent English words that have come to be this way in the past couple decades, e.g. bad. – Kosmonaut Jun 14 '11 at 15:38
  • 1
    I don't think I'd categorize "egregious" as such. It seems to me to be a garden variety intensifier that usually carries negative connotations but can be pressed into service for the opposite effect. – phenry Jun 14 '11 at 16:53
  • 1
    English is sophisticated that way. – Hot Licks Dec 14 '15 at 1:30
  • "Egregious" seems to mostly have been used to express a negative connotation. I don't think it classifies so much in this sense. – xji Feb 9 '17 at 23:00
up vote 11 down vote accepted

If you start with wikipedia

An auto-antonym (sometimes spelled autantonym), or contranym (originally spelled contronym), is a word with a homograph (a word of the same spelling) that is also an antonym (a word with the opposite meaning). Variant names include antagonym, Janus word (after the Roman god), enantiodrome, and self-antonym. It is a word with multiple meanings, one of which is defined as the reverse of one of its other meanings.

it already puts you on the right path and mentions one of the word and one of the ways that can make such words

Some pairs of contronyms are true homographs, i.e., distinct words with different etymology which happen to have the same form. For instance cleave "separate" is from Old English clēofen, while cleave "adhere" is from Old English cleofian, which was pronounced differently. This is related to false friends, but false friends do not necessarily contradict.

In other words, for example literally, the two meanings developed from the overuse of the word as hyperbole.

  • That seems such an incredible coincidence, though, doesn't it? That two words with completely distinct (actually opposite) meanings, and distinct pronounciations should converge over time. – Seamus Jun 14 '11 at 15:36
  • @Seamus, actually no, the coincidence is not so huge I believe. For example when one language influenced another through imperfect means (sporadic conquest or access to incomplete libraries, etc) it is common to change the meaning of the words due to local meaning. I would not rule out simplifying or simply wrongly hearing words, or merging similar words, etc.. – Unreason Jun 14 '11 at 15:51

It's probably fair to say that the processes of language change aren't so well understood as other areas of linguistics. But, the following probably have an influence:

  • certain words tend to get used alongside certain others as fixed colloquations
  • as particular colloquations become common, the meaning of the individual words probably gets "spread" over the colloquation as a whole, and speakers don't necessarily process the individual words with their individual meanings, thus allowing the part of the overall meaning that's attributed to each individual word of the colloquation to "shift" slightly
  • colloquations may be created and then shortened so that you accidentally end up with cases such as "terrific" (I suspect the process was something like "terrific" > "terrifically bad" > "terrifically good" > "terrific")
  • in longer colloquations, you can replace one word with another of quite different meaning, and people may still interpret the overall colloqation has having a similar meaning because of this effect, e.g. you'll obesrve speakers using "It leaves much to be desired" vs "It leaves little to be desired" more or less interchangeably
  • languages are "complex systems", and the frequency that words are used in different colloquations/usages can vary quite chaotically; meanwhile, if a word is used with a particular meaning only in a handful of colloquations, it is probably liable for those "minority" colloquations to be re-interpreted to fit the majority usage of a given word (this appears to have happned e.g. with "begs the question", which now tends to mean "requires the question to be asked" rather than "ignores the question", or "more haste less speed", where "speed" tends to have a more literal interpretation than its original meaning that was probably closer to "success")
  • specifically with the case of words taking on their opposite meanings, a possible influence is that speakers appear to be quite bad at processing negatives. For example, speakers are liable to process pairs such as , or "No accident is too severe to ignore" vs "No accident is too trivial to ignore" as effectively meaning the same thing, even though "severe" and "trivial" have pretty much opposite meanings (this is similar to the "much to be desired" example, but specifically having a negative in the sentence appears to enhance the effect)
  • there are probably cases where speakers deliberately pick the antonym of a word and use it with the same meaning ("you loser" > "you winner"), or deliberately choose a negative expression and actively use it with a positive meaning ("that's totally sick, man!")
  • 3
    "begs the question" originally meant "presupposes an answer to the question" rather than "ignores the question", right? – Seamus Jun 14 '11 at 15:59
  • 2
    Seamus -- actually, I'm not sure whether that meaning or the "ignore" meaning came first, now you mention it. But I think nowadays, people tend to use it to mean "raises the question" rather than "ignores" or "assumes already answered". – Neil Coffey Jun 14 '11 at 16:18
  • Language log seem to be suggesting that at least as far back as Aristotle it was a term for a kind of circular reasoning – Seamus Jun 14 '11 at 16:29
  • The Language Log article suggests another (more minor) item for my list: uses that start as calques/ad hoc translations, and then the original word is forgotten. The article also demonstrates the fallacy of the "language shouldn't change argument": if you were forced to worry in that level of detail about every single word, you'd never actually say anything... – Neil Coffey Jun 14 '11 at 17:24

The evolution of language is an inherently human process. Words are coined to mean one thing, but can change based on various cultural factors, including geographic, demographic, financial, and of course historical.

For example, there is a verb used in political discussions, "to table". Common usage: "The bill to increase the police budget was tabled". What's odd is that this word means two different things depending on which side of the pond you're on: in America, the term is shorthand for "to lay on the table"; that is, to dismiss from or halt discussion. However, in Britain, the term is shorthand for "to bring to the table", that is, to introduce for discussion". Thus, "tabling" a bill is the start of the process in Britain, but a bad end in America.

This illustrates just one way the same word can evolve to have two different meanings; the shortening of phrases with opposite meanings to the same word or term. These two meanings can then coincide in the same culture if users of the two meanings become culturally commingled.

More to your case in point, let's take the word "cleave". In modern usage, the word has the connotation of a split. However, the word's original meaning is closer to "to penetrate into". Consider an axe splitting wood. The act of "cleaving" the wood happens as the axe enters into the log. If the axe continues all the way through, the log is "cleft in twain"; split in two, hence that connotation. However, if it doesn't, the log is still cleft, but as anyone who's chopped firewood before knows, an axe buried halfway into a log is difficult to remove. A similar connotation evolves from the obvious sexual underpinnings of the biblical phrase "to cleave to one another [and become one flesh]"; the "penetration" causes a bond between the two people, making them one (in many ways). So, the actual historical meaning of "cleave" thus can result in two different colloquial meanings depending on the pictured result.

As far as "egregious", I'm not sure that qualifies. The word has always (as far as Webster's concerned) been synonymous with "extraordinary". This can be good or bad depending on the subject of the adjective. The usual usage is with the connotation that what is egregious is a bad thing: "an egregious series of errors". However, the term can be used positively: "an egregious donation". The synonym to "extraordinary" is obvious if you swap one for the other, and in British English both terms are used as superlatives in both directions, good and bad. In American English, "extraordinary" is normally used to emphasize good, while "egregious" emphasizes bad.

  • 2
    Actually, as Unreason said, the two meanings of "cleave" are actually from different roots. But generally a good answer. – Colin Fine Jun 14 '11 at 16:42
  • "To lay upon the table" is also used in Parliament in the UK. It is the way in which documents are formally presented to Parliament, and so, as you say, represents the start of their legislative journey, rather than the end. For example, see the appendix in the most recent House of Commons Votes & Proceedings. – Steve Melnikoff Jun 14 '11 at 22:27

People do sometimes use a word in a sense that could be construed as strongly (and perhaps diametrically) opposed to its usual sense (at a particular time), through simple misunderstanding of how the term is normally used. Sometime later, if that usage catches on, dictionaries that report the way people actually use the word begin including the opposed meaning as one possible legitimate definition of the word.

Thus, for example, Merriam-Webster's Collegiate Dictionary, Eleventh Edition (2003) includes this definition of infer (a word whose earliest retained definitions are "to derive as a conclusion from facts or premises" and "to guess or surmise"):


But these words are equivalent to "express indirectly" which is the most common sense in which modern English speakers use the term imply; and imply occurs at the opposite end of a traditional imply–infer pipeline from the end at which infer occurs. According to the older understanding, a hinter implies some piece of intelligence, and a hintee infers it. According to the newish definition of infer (which goes back, in one form or another to the Seventh Collegiate of 1963, before which time Webster's labeled the "imply" meaning of the term "Erroneous"), however, inferring can take place at both ends of the pipeline.

A similar situation might have arisen in the case of inflammable, which means "capable of being easily ignited" but which early-20th-century fire marshals were concerned might be misunderstood to mean "not capable of being easily ignited"—so concerned, indeed, that they introduced flammable as a replacement word to take over the easy-ignition definition from inflammable. You might suppose that this development cleared the way for popular usage of inflammable to mean "not flammable"—the opposite of its original meaning—but Merriam-Webster's doesn't acknowledge that any such altered usage has occurred.

I don't have any sources to cite on this one, since I'm merely riding on my own observation. However, from how I understand it, this process usually takes place when one context of the word gets stuck in social memory. For example, I could say of someone, "Adam Rackis has made an egregious mistake," meaning that this is a mistake so awful that no one else would ever have made it. What sticks in people's minds, though, is the fact that I am communicating that Adam has made an horrific mistake, not the actual meaning of the word. Over time, as the connotation is remembered and the definition is not, the word "egregious" becomes associated with something terrible rather than something unusual. Thus, this kind of evolution of language is based on ignorance.

  • I like the ignorance explanation, like the use of "literally" to mean (effectively) "figuratively" as in Unreason's answer... – Seamus Jun 14 '11 at 15:35
  • With your (jump to) conclusion you ignore the fact that all words were once metaphors and that the usage can organically change language in ways that have nothing to do with ignorance or knowledge. – Unreason Jun 14 '11 at 15:49
  • Yes, all words were once metaphors. And all lightning storms were once the wrath of the great god Tengri. – Andrew Jun 14 '11 at 16:03
  • @Unreason language is conventional sure. That doesn't mean people can't be ignorant of the proper conventions. And the conventions change over time, sure. But that doesn't mean that words don't have "correct" meanings. – Seamus Jun 14 '11 at 16:15
  • @Andrew, Seamus: here are two quotes I like as much for their precision and erudition as for their style: 1) Emerson said that language is fossil poetry. As confirmation of this dictum, we need only remember that all abstract words are, in effect, metaphors, including the word metaphor, which in Greek means "transfer." —Jorge Luis Borges, Argentinian poet, short-story writer, and essayist, Atlas, 1985 2) After all, what are words? Words are symbols for shared memories. Jorge Luis Borges, This Craft of Verse – Unreason Jun 14 '11 at 16:33

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.